Especially among girls, the adolescent transition is associated with dramatic increases in the prevalence of suicidal ideation, and several forms of self-injury, including non-suicidal self-injury (NSSI; i.e., self-mutilation), suicidal gestures, threats, and attempts. This study proposes, and will test a theoretical model suggesting that biological and cognitive responses to social stressors explain the association between psychopathology (i.e., depression) and self-injury (i.e., mediation) and that specific interactions between stress responses (i.e., moderation) will help identify which girls with psychopathology are most likely to engage in self-injury longitudinally. Specific combinations of cognitive and biological stress responses are proposed to uniquely identify risks for different types of self-injury/suicidality (i.e., NSSI vs. suicidal ideation). Moreover, this study will examine trajectories of, and associations among self-injury constructs, addressing several limitations of past work. The aims of this research thus address several of the goals outlined in PA # 07-079, Research on the Reduction and Prevention of Suicidality as well as many of the goals articulated the NIMH Strategic Plan (NIMH, 2008) and the NAMHC Workgroup report on Transformative Neurodevelopmental Research (NAMHC, 2008). This study will use an innovative, lab-based methodological paradigm to examine cognitive (i.e., attributions, social problem solving) and biological (i.e., neuroendocrine, cardiovascular) responses to an in vivo social stressor. Participants will include 250 female adolescents from both outpatient and inpatient clinically-referred samples. Data will be collected from multiple informants (adolescents, parents) and multiple sources (observational methods, structured interviews, questionnaires, biological assays). It is expected that observed stress responses in the lab will interact with the experience of actual social stress measured during follow-up to predict self-injury trajectories over an 18 month interval. In other words, this study will address long-standing, but under-explored questions regarding why and how psychological symptoms, and/or the experience of stress, are associated with self-injurious behaviors.